Emanuele Trucco

Recent clinical research has highlighted important links between a number of diseases and the tortuosity of curvilinear anatomical structures like corneal nerve fibres, suggesting that tortuosity changes might detect early stages of specific conditions. Currently, clinical studies are mainly based on subjective, visual assessment, with limited repeatability and inter-observer agreement. To address these problems, we propose a fully automated framework for image-level tortuosity estimation, consisting of a hybrid segmentation method and a highly adaptable, definition-free tortuosity estimation algorithm. The former combines an appearance model, based on a Scale and Curvature-Invariant Ridge Detector (SCIRD), with a context model, including multi-range learned context filters. The latter is based on a novel tortuosity estimation paradigm in which discriminative, multi-scale features can be automatically learned for specific anatomical objects and diseases. Experimental results on 140 ...

To assess the performance of a novel system for automated tortuosity estimation and interpretation. A supervised strategy (driven by observers' grading) was employed to automatically identify the combination of tortuosity measures (i.e., tortuosity representation) leading to the best agreement with the observers. We investigated 18 tortuosity measures including curvature and density of inflection points, computed at multiple spatial scales. To leverage tortuosity interpretation, we propose the tortuosity plane (TP) onto which each image is mapped. Experiments were carried out on 140 images of subbasal nerve plexus of the central cornea, covering four levels of tortuosity. Three experienced observers graded each image independently. The best tortuosity representation was the combination of mean curvature at spatial scales 2 and 5. These tortuosity measures were the axes of the proposed TP (interpretation). The system for tortuosity estimation revealed strong agreement with the observers on a global and per-level basis. The agreement with each observer (Spearman's correlation) was statistically significant (αs = 0.05, P < 0.0001) and higher than that of at least one of the other observers in two out of three cases (ρOUR = 0.7594 versus ρObs3 = 0.7225; ρOUR = 0.8880 versus ρObs1 = 0.8017, ρObs3 = 0.7315). Based on paired-sample t-tests, these improvements were significant (P < 0.001). Our automated system stratifies images by four tortuosity levels (discrete scale) matching or exceeding the accuracy of experienced observers. Of importance, the TP allows the assessment of tortuosity on a two-dimensional continuous scale, thus leading to a finer discrimination among images.

We report a novel application of polarization based vision addressing the robustness of laser triangulation range sensors. Such sensors are based on the accurate detection of a pattern of laser light projected onto a scene, usually a point or line. Typical problems arise with highly specularly reflective surfaces, which can generate visible reflections of the light in various parts of

... looking sonar presents itself as a versatile option and has been integrated into our framework. ... image sequence (see [15] for a survey of robust methods in computer vision, and [16 ... A video sequence acquired by a single camera in unknown motion contains complete information ...

... forwiss. de/" munkel t Dr. Thomas Munsterer V1TRONIC Dr.-Ing. Stein Bildverarbeitungssysteme GmbH Hasengartenstr. 14a, D-65189 Wiesbaden, Germany Phone:+ 49-611-7152-38, tm@ vitronic. de Prof. ... uni-heidelberg. de http://www. i up. uni-heidel berg. de/urmel/atmos. ...

We propose a new approach to corneal nerve fibre centreline detection for in vivo confocal microscopy images. Relying on a combination of efficient hand-crafted features and learned filters, our method offers an excellent compromise between accuracy and running time. Unlike previous solutions using sparse coding to learn small filter banks, we employ K-means to efficiently learn the high amount of filters needed to cope with the multiple challenges involved, e.g., low contrast and resolution, non-uniform illumination, tortuosity and confounding non-target structures. The use of K-means for dictionary learning allows us to learn banks of 100 filters in less than 30 seconds compared to several days needed when using sparse coding. Experimental results using a dataset including 100 images show that our approach outperforms significantly state-of-the-art methods in terms of precision-recall curves.

Heriot Watt University, Riccarton, Edinburgh, EH14 4AS, UK e.trucco@hw.ac.uk ... This paper present a novel, rohust apprwach to thc automatic location ofthe optic disk in retinal (fundus) images. Instcad of generating a single, high-confidence optic disk candidate, we gcncratc ...

Considers two well-known pattern recognition techniques using texture analysis. The first is the co-occurrence matrix method which relies on statistics and the second is the Kohonen map which comes from the artificial neural networks domain. Both methods are used as feature extraction methods. The extracted feature vectors are fed to a second Kohonen map used as classifier. The authors report briefly some results of their experimental assessment of the merit of each technique when applied to the problem of classifying the seabed from sequences of real images

Sarcopenia in older people is a major health issue and its early detection could help target interventions and improve health. Evidence suggests that poor muscle mass is associated with greater arterial stiffness and cardiovascular risk. Arterial stiffness in turn is associated with smaller retinal artery width. This study examined the association of muscle mass in older people with retinal vascular width, a non-invasive measure of vascular function. Participants >65 years were recruited to a cross-sectional study. Inability to walk independently; diabetes mellitus; stroke (within 6 months), severe macular degeneration, glaucoma, retinal dystrophy; advanced cataract. Digital Retinal images of both eyes were analysed using the VAMPIRE software suite. Central Retinal Artery and Vein Equivalents (CRVE and CRAE) were measured. Body composition was measured using Dual Energy X ray Absorptimetry (DXA). Appendicular Skeletal Muscle Mass/Height(2) was calculated. Physical function was measured: 6-min walk distance, Short Physical performance battery, handgrip strength and quadriceps strength. 79 participants with mean age 72 (SD 6) years were recruited. 44% were female. Digital Retinal images of sufficient quality for measuring CRAE and CRVE were available for 51/75 (68%) of participants. Regression analysis showed significant association between larger ASMM/H(2) and smaller CRAE (β=-0.20, p=0.001) and CRVE (β=-0.12, p=0.05). Handgrip strength, body mass index and sex combined with CRAE explained 88% and with CRVE explained 86% of the variance in ASMM/H(2). Larger muscle mass was significantly associated with smaller retinal artery size in older people. This unexpected finding needs further investigation.

In this paper, we present a new adaptive approach to multi-view markerless articulated human body pose estimation from multi-view video sequences, using Particle Swarm Optimisation (PSO). We address the computational complexity of the recently developed hierarchical PSO (HPSO) approach, which successfully estimated a wide range of different motion with a fixed set of parameters, but incurred an unnecessary overhead in

ABSTRACT In this paper, we propose a novel face authentication approach based on affine scale invariant feature transform (ASIFT) and structural similarity (SSIM). The ASIFT descriptor defines key points which are used to match the gallery and probe face images. The matched pairs of key points are filtered based on the location of points in the gallery face image. Then the similarity between sub-images at a preserved pair of matched points is measured by Structural Similarity (SSIM). A mean SSIM (MSSIM) at all pairs of points is computed for authentication. The proposed approach is tested on FERET, CMU-PIE and AR databases with only one image for enrollment. Comparative results on the AR database show that our approach outperforms state-of-the-art approaches.

... 1345 Page 2. in which z = [dll d12 d21 d22 dl d2IT contains the un-known motion parameters, and T a = --7 ~t [ UI ~ UI,, VI^ VI,, I, I,,] , W with UVI: UVI,I,, V~I,,I, ' U”: u21uI, UVI: uvIuI, uvIuI,, UVI; v21uIu U”; 1 = [ u21uI,, U”,” UVI,I, uv1; ] UI: uIuI,, VI: VT = [ uIuIu UI,” VI& ...

Abstract We present an uncalibrated vision system recovering 3D reconstructions of underwater scenes using no camera calibration at all. The only information available is a set of point correspondences plus some a-priori knowledge on the scene structure. The system includes three main components: a robust tracker, which targets optimal features for video tracking, and implements a completely automatic mechanism for rejecting unstable image features; a module computing a projective reconstruction of the scene from the set of ...

We report a novel algorithm to locate vascular leakage and ischemia in retinal angiographic image sequences leveraging contextual knowledge of co-occurring pathologies. The key contributions are the use of spatio-temporal features exploiting the evolution of intensity levels over the sequence and contextual knowledge to detect ischemia. The specific nature of these diseased regions is determined using an AdaBoost learning algorithm. Training was performed with a varied set of 16 ground-truth image sequences, and testing on unseen images. The images used were acquired with an Optos ultrawide-field scanning laser ophthalmoscope. Evaluation against manual annotations demonstrates successful location of 93% of leakage regions and 70% of ischemic regions.

This paper presents an analytical study and several practical results for computing stable reconstructions with uncalibrated, motion-based inspection and conveyor-belt installations. We achieve metric recon- struction with a simple, efficient algorithm and two nonlinear con- straints expressing knowledge readily accessible in a real setup. We analyse the stability and accuracy of reconstruction with respect to the system's mathematical structure, pixelisation, image noise, and constraint values. Extensive experiments with simulated and real data have confirmed our analysis in full, and one example is illustrated here.

An investigation was undertaken into the aptitude of Genetic Programming for solving the problem of visibility spaces for Sensor Planning in a Machine Vision system. A number of approaches, varying the tness function and the virtual machine on which the genetic programs run, are tried out, and it is concluded that this problem is not one easily amenable to solution by means of Genetic Programming. An analysis is undertaken of the reasons that this might be the case.

... 1985. [7] YH Liu and S. Arimoto, “Path planning using a tangent graph for mobile robots among polygonal and curved obstacles,” Int. J. Robot. ... 3075–3080. Scheduling Messages on Controller Area Network for Real-Time CIM Applications Khawar M. Zuberi and Kang G. Shin ...

Division of Informatics Forrest Hill & 80 South Bridge. Publication Details: Book: Proceedings of the 11th Brazilian Symposium on Artificial Intelligence. Editors: Carvallis, Tarcisco Pequeno an; Carvallis, Tarcisco Pequeno and Fernando. Book#: MISSING. Publisher: MISSING. Volume: MISSING. Pages: MISSING. Paper in Conference Proceedings #732. Title: Estracting Prolog Programming Techniques. Authors: Vasconcelos,W. Date: 1994. Presented: Keywords: ...